gavindiaz/mymt5opp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add requirements.txt and rewrite README for handover

Browse Source
This commit is contained in:
gavindiaz
2026-06-27 01:05:34 +08:00
parent e99c14e907
commit 51f9a6b98d
5 changed files with 395 additions and 295 deletions
Download Patch File Download Diff File Expand all files Collapse all files
COMMIT_MSG_2.txt
-10
View File
@@ -1,10 +0,0 @@
warmup 修复后重跑 Optuna study + 重生所有产物
- 重跑 500 trials Optuna study (warmup 修复版)
- 新 finalist 集合: #1 trial #491 / #2 trial #297 / #3 trial #492
- finalist #1 IS net $263k (vs 旧 $29k, warmup 修复后搜索空间扩大)
- 重生 finalists JSON / features parquet / dashboard HTML
- 三个独立 registry 条目 (f1_t491 / f2_t297 / f3_t492)
- registry 文件名含 finalist index + trial number 避免覆盖
- registry 状态: "已记录 (无 MT5 验证)" - 待 Phase 7 MT5 验证
- 删除旧 registry 条目 (trial #324 已被新 study 取代)
README.md
+372 -88
View File
@@ -1,117 +1,401 @@
# Backtesting + Optuna + MT5-Verification Stack — Knowledge Base # Backtesting + Optuna + MT5 Stack
A blueprint for building a **personal trading-strategy research lab**: a fast Python 一个把 **MetaTrader 5 Strategy Tester 当作"金标准"**、用 **Python 镜像引擎**做高速贝叶斯
backtesting engine, a Bayesian parameter optimizer (Optuna), and an automated bridge to the 搜索的个人量化策略研究实验室。当前已对 **GoldScalperPro**（XAUUSD M5 trailing/BE EA）
**MetaTrader 5 Strategy Tester** that cross-checks every result against the real terminal. 跑通完整闭环：**假设 → 搜索 → MT5 验证 → 入注册表**。
This is a **knowledge base, not a code drop.** It describes the *architecture, algorithms, rules, > 接手者请按本文件的 §2 → §3 → §7 顺序读完即可上手。详细的架构 / 坑 / 扩展指南
and tooling* so you can build your own version from scratch — with **your own** Expert Advisors, > 在 [PROJECT_GUIDE.md](PROJECT_GUIDE.md) 和 `01-08-*.md` 里。
**your own** strategies, and **your own** presets. There are no ready-made engines or strategy
files here on purpose: you supply those from your own MQL5 bots (ideally open-source EAs you own
or have the right to use).
--- ---
## What this stack does for you ## 1. 项目一句话
You have MQL5 Expert Advisors and a pile of ideas to test. The MetaTrader Strategy Tester is
accurate but **slow** — a single multi-year backtest with a real-tick model can take 10–30 minutes,
and an exhaustive optimization can run for days. That kills iteration speed.
This stack solves it with a **two-tier model**:
1. **Tier 1 — a fast Python "mirror engine"** that reproduces your EA's trade logic bar-by-bar over
pre-downloaded historical data. It is an *approximation* (more on fidelity below), but it runs a
full multi-year backtest in **seconds**, so you can rank thousands of parameter combinations with
Optuna in the time MT5 would run a handful.
2. **Tier 2 — MetaTrader 5 as the gold standard.** Only the **finalists** from the Python search are
compiled and run in the real Strategy Tester. The Python number gets you the *shortlist*; the MT5
number is what you trust for anything that goes live.
``` ```
Idea ─► Python mirror engine ─► Optuna search (thousands of trials, fast) Idea ─► Python mirror engine ─► Optuna search (hundreds of trials, seconds each)
│ │
▼ ▼
2–3 diverse finalists 2–3 diverse finalists
│ │
▼ ▼
MetaTrader 5 Strategy Tester (gold standard) MetaTrader 5 Strategy Tester (gold standard)
│ │
▼ ▼
Python-vs-MT5 comparison table ─► keep / discard / iterate Python-vs-MT5 gap table ─► registry/ (append-only)
``` ```
The whole point is **disciplined isolation**: the engine is frozen and trusted, instruments are - **Tier 1（Python 镜像引擎）**：把 EA 的填单 / 出场逻辑 bar-by-bar 重写，跑在
described by data not code, every hypothesis is an isolated experiment, and only triple-checked Parquet 历史数据上，2 年回测几秒钟。
results are promoted. That discipline is what keeps a research lab from rotting into a pile of - **Tier 2（MT5 Strategy Tester）**：只对 finalist 跑真实 tester。**MT5 数字才是
one-off scripts that nobody can reproduce. live 决策依据**；Python 数字只负责排序。
- **trailing/BE EA 必须 M1 tick-level 模拟**——bar-level 会产生 −40% 到 −50%
的隐藏 gap（详见 [PROJECT_GUIDE.md §1.3](PROJECT_GUIDE.md)）。
--- ---
## Who this is for ## 2. 快速开始（接手者从这里读起）
- You run **MetaTrader 5** and write or use **MQL5** Expert Advisors. ### 2.1 环境要求
- You want to test and optimize strategies **much faster** than the MT5 optimizer allows.
- You are comfortable with **Python** (intermediate) and the command line.
- You can run **MT5 on Windows** — either on the same machine (simplest) or on a separate
Windows box/VPS that your Python machine talks to.
You do **not** need to be a quant. The hard parts (engine fidelity, Bayesian search, robustness - **Windows**（MetaTrader5 Python 包只在 Windows 上装得了）
testing) are explained from first principles. - **Python 3.12+**（开发用 3.12.10）
- **MetaTrader 5 终端**已安装并登录一个 demo 账户
- **Git**（commit 历史就是项目文档的一部分）
### 2.2 一次性准备
```powershell
# 1. clone 后建 venv 装依赖
python -m venv .venv
.\.venv\Scripts\activate
pip install -r requirements.txt
# 2. 在根目录建 .env（gitignored，凭证不入库）
# MT5_DEMO_LOGIN=52845377
# MT5_DEMO_PASSWORD=...
# MT5_DEMO_SERVER=ICMarketsSC-Demo
# MT5_TERMINAL_PATH=C:\Program Files\MetaTrader 5 IC Markets Global\terminal64.exe
# 3. 打开 MT5 终端、登录 demo、确认 XAUUSD 历史已下载（M5 + M1）
python scripts/download_xauusd_history.py # → data/XAUUSD_M5_*.parquet
python scripts/download_xauusd_m1.py # → data/XAUUSD_M1_*.parquet（trailing/BE EA 必需）
# 4. 查 symbol 规格（确认 tick_value / contract_size / lot_step）
python scripts/query_xauusd_spec.py
```
历史 parquet 在 `.gitignore` 里（`data/` 整个忽略），所以**接手后必须重跑下载脚本**，
否则后续 Optuna / 评估脚本会 `sys.exit("missing M5 data")`。
### 2.3 5 分钟跑通一遍
```powershell
# A. 烟雾测试：30 trials，IS H1 2025，<2 分钟验证全链路通
python scripts/optimize.py --smoke
# B. 全量搜索：500 trials，IS = 2025 全年，~10–20 分钟
python scripts/optimize.py --trials 500
# 断点续跑（load_if_exists=True），中途 Ctrl+C 不丢
# 产物：studies/optuna/gold_scalper_pro_is2025.db + .log
# C. 用 finalist 参数重跑 Python（warmup 修复版），落 JSON
python scripts/reeval_finalist_forward.py
# 产物：studies/finalists/gold_scalper_pro_is2025-2026.json
# D. 生成中文 Optuna 交互式 dashboard（HTML）
python scripts/build_optuna_dashboard.py
# 产物：reports/optuna_dashboard_gold_scalper_pro_is2025.html
# E. 生成 ML 特征数据集（trade-level + trial-level）
python scripts/build_feature_datasets.py
# 产物：studies/features/{trade,trial}_features_*.parquet + .csv
# F. 自动生成 registry markdown 条目（每 finalist 一份）
python scripts/build_registry_entry.py
# 产物：registry/gold_scalper_pro_xauusd_*_f{idx}_t{trial}.md
```
**MT5 验证是手动步骤**（Strategy Tester 跑完导 HTML 到 `reports/`），跑完后用：
```powershell
python scripts/build_registry_entry.py `
--mt5-is-html reports\IS-ReportTester-52845377.html `
--mt5-oos-html reports\OOS-ReportTester-52845377.html
```
补齐 registry 条目里的 MT5 指标和 gap 表。
--- ---
## How to read this KB ## 3. 目录结构
Start with `CLAUDE.md` if you plan to use an AI coding assistant (Claude Code, Cursor, etc.) to ```
build this with you — it is an **adaptive setup playbook** that profiles *your* machine and OS and backtesting-optuna-mt5-stack/
walks the install from zero. Otherwise read the numbered docs in order: ├── README.md ← 本文件，项目入口
├── PROJECT_GUIDE.md ← 项目实例说明（架构+坑+使用+扩展，10 节）
| # | Doc | What you get | ├── CLAUDE.md ← 给 AI 助手的分阶段搭装 playbook
|---|-----|--------------| ├── requirements.txt ← Python 依赖清单
| — | [`CLAUDE.md`](CLAUDE.md) | Adaptive AI-assistant playbook: profiles your device, drives install from scratch. Doubles as `AGENTS.md`. | ├── .gitignore ← data/ + .env 忽略；studies/*.db + reports/*.html 入库
| 01 | [`01-stack-and-install.md`](01-stack-and-install.md) | The exact tech stack, every library, where to get it, and how to install — per OS. | │
| 02 | [`02-architecture.md`](02-architecture.md) | The layered architecture and data flow. The mental model for everything else. | ├── 01-stack-and-install.md ← 知识库 §1：技术栈 + 每个系统的安装命令
| 03 | [`03-engine-design.md`](03-engine-design.md) | How to design your own bar-by-bar engine. Grid logic used as the worked example. Fidelity vs MT5. | ├── 02-architecture.md ← 知识库 §2：单向依赖分层 + 数据流
| 04 | [`04-isolation-rules.md`](04-isolation-rules.md) | The isolation discipline: frozen engines, forks, separated instruments/strategies, curated registry. | ├── 03-engine-design.md ← 知识库 §3：bar-by-bar engine + intra-bar 4-sub-tick
| 05 | [`05-config-and-inputs.md`](05-config-and-inputs.md) | How test inputs are kept *separate*: instrument config, parameter space, the pre-run wizard, lot/money mode. | ├── 04-isolation-rules.md ← 知识库 §4：8 条隔离铁律
| 06 | [`06-optimization-and-robustness.md`](06-optimization-and-robustness.md) | Optuna objective design, constraints, diverse top-N selection, and anti-overfit robustness layers. | ├── 05-config-and-inputs.md ← 知识库 §5：4 个声明输入源
| 07 | [`07-mt5-bridge.md`](07-mt5-bridge.md) | Connecting to MT5: compiling EAs, auto-running the tester, parsing reports, comparing Python vs MT5. Local-Windows and remote variants. | ├── 06-optimization-and-robustness.md ← 知识库 §6：Optuna objective + 反过拟合
| 08 | [`08-workflow-cycle.md`](08-workflow-cycle.md) | The full repeatable cycle: hypothesis → scaffold → stats → optimize → verify → promote. | ├── 07-mt5-bridge.md ← 知识库 §7：编译 EA / .set / .ini / 解析报告
├── 08-workflow-cycle.md ← 知识库 §8：8 步可重复循环
│
├── GoldScalperPro.mq5 ← EA 源码（MQL5）
├── GoldScalperPro.ex5 ← EA 编译产物
│
├── shared/ ← 9 个单向依赖层（低调高、高不知低）
│ ├── core/ ← engine（冻结）+ metrics
│ ├── data/ ← Parquet loaders + MT5 HTML parser
│ ├── gates/ ← 入场过滤 mask（regime / 时段 / exhaustion）
│ ├── indicators/ ← 纯函数：RSI / ATR / EMA / SMA
│ ├── instruments/ ← per-symbol config（tick_value / spread / swap）
│ ├── mt5_pipeline/ ← .set / .ini / compile / runner / compare
│ ├── optimizer/ ← objective / search_space / diverse top-N selector
│ ├── robustness/ ← 反过拟合分析（只读 over results）
│ ├── wizard/ ← 运行时 Q&A → wizard-answers.yaml
│ └── config.py ← .env 加载（含 plain fallback）
│
├── strategies/
│ └── gold_scalper_pro/ ← glue：data → signals → stops → engine → metrics
│ ├── instruments.py ← XAUUSD_REAL config 对象
│ ├── params.py
│ ├── scalper_engine.py ← ScalperEngine（engine.run 的 wrapper）
│ ├── search_space.py ← 19 个可调参数 + 13 个冻结基线 + 约束
│ ├── set_mappings.py ← Python 参数名 ↔ MQL5 input 名映射
│ ├── signals.py ← build_signals（EMA cross + RSI + ATR）
│ └── wizard_questions.py
│
├── scripts/ ← 一次性可执行脚本（按用途分组见 §5）
│ ├── optimize.py ← Phase 6 主入口（smoke / full）
│ ├── reeval_finalist_forward.py ← finalist 参数 + warmup 修复版重跑
│ ├── build_optuna_dashboard.py ← 中文 plotly 交互式 HTML
│ ├── build_feature_datasets.py ← ML 特征 parquet
│ ├── build_registry_entry.py ← 自动生成 registry markdown（命令驱动，禁手写）
│ ├── walk_forward.py
│ ├── prepare_mt5_verify.py ← 生成 .set / .ini 给 MT5 tester
│ ├── verify_mt5.py
│ ├── compare_finalist.py ← Python vs MT5 差距表
│ ├── diag_*.py ← 9 个诊断脚本（ATR / 时区 / 引擎 trace / 信号）
│ └── download_xauusd_*.py ← 历史 M5 / M1 数据下载
│
├── studies/ ← Optuna 研究产物（committed）
│ ├── optuna/ ← SQLite DB + 运行日志（resumable）
│ │ ├── gold_scalper_pro_is2025.db
│ │ ├── gold_scalper_pro_is2025.db.before-warmup-fix.bak ← 旧 study 备份
│ │ └── gold_scalper_pro_is2025.log
│ ├── finalists/ ← 3 个 finalist 的 IS/OOS 指标 JSON
│ │ └── gold_scalper_pro_is2025-2026.json
│ └── features/ ← ML 特征数据集（parquet + csv 副本）
│ ├── trade_features_*.parquet ← trade-level 35 列
│ └── trial_features_*.parquet ← trial-level 聚合
│
├── reports/ ← MT5 + dashboard HTML（committed）
│ ├── IS-ReportTester-52845377.html
│ ├── OOS-ReportTester-52845377.html
│ └── optuna_dashboard_gold_scalper_pro_is2025.html
│
├── registry/ ← 已批准 finalist（append-only，真相源）
│ ├── gold_scalper_pro_xauusd_*_f1_t491.md ← finalist #1（trial #491）
│ ├── gold_scalper_pro_xauusd_*_f2_t297.md ← finalist #2（trial #297）
│ └── gold_scalper_pro_xauusd_*_f3_t492.md ← finalist #3（trial #492）
│
└── data/ ← gitignored，历史 parquet（接手后需重下）
├── XAUUSD_M5_2024-06-26_2026-06-26.parquet
└── XAUUSD_M1_2024-06-26_2026-06-26.parquet
```
--- ---
## The 30-second mental model ## 4. 核心数据流（一个 iteration 的闭环）
- **The engine knows nothing about your strategy.** It takes bars + entry signals + stop/target ```
prices and simulates fills. All strategy math lives in *caller* code. (Doc 02–03.) ┌──────────────────────────────────────────────────────┐
- **The engine is frozen.** You never edit a validated engine to test an idea — you fork it, │ scripts/optimize.py │
prove the fork reproduces the original 1:1 with the change off, then test. (Doc 04.) │ load_m5() ─► slice_window(IS_START, IS_END) │
- **Instruments are data, not code branches.** Tick value, spread model, swap, lot steps — all in a │ load_m1() ─► slice_window(IS_START, IS_END) │
per-symbol config object. The engine reads everything from it. (Doc 05.) │ full_m5 ───────────────────┐ (warmup, signals_full_bars)│
- **Search inputs are declared, not scattered.** Every tunable parameter, its range, and its │ ▼ │
constraints live in one declared search space; one wizard captures the run settings; one YAML │ ObjectiveConfig(bars_is, m1_is, signals_full_bars=full_m5)│
records the answers so any run is reproducible. (Doc 05–06.) │ │ │
- **Python ranks, MT5 decides.** Fast Python search produces a shortlist; MT5 produces the trusted │ ▼ │
number; a comparison table is saved for every finalist. (Doc 03, 06, 07.) │ build_objective(cfg) ─► trial loop │
│ │ │
│ ▼ │
│ build_signals(full_m5) ─► slice to IS │
│ engine_kwargs_from_params(params) │
│ ScalperEngine.run(bars, signals, sl, tp, m1_bars) │
│ │ │
│ ▼ │
│ compute_metrics + Constraints gate │
│ study.optimize() ─► trials 0..499 │
└──────────────────────────────────────────────────────┘
│
▼
┌──────────────────────────────────────────────────────┐
│ select_diverse_topn(study, n=3) │
│ └─► 3 个 diverse finalist（greedy max-distance） │
└──────────────────────────────────────────────────────┘
│
▼
┌──────────────────────────────────────────────────────┐
│ scripts/reeval_finalist_forward.py │
│ 用 finalist params 重跑 Python（warmup 修复版） │
│ ─► studies/finalists/gold_scalper_pro_is2025-2026.json│
└──────────────────────────────────────────────────────┘
│
▼
┌──────────────────────────────────────────────────────┐
│ MT5 Strategy Tester（手动） │
│ prepare_mt5_verify.py 生成 .set + .ini │
│ Forward mode: IS=2025, OOS=2026 H1 │
│ 导 HTML 报告到 reports/ │
└──────────────────────────────────────────────────────┘
│
▼
┌──────────────────────────────────────────────────────┐
│ scripts/build_registry_entry.py │
│ finalist JSON + Optuna study + MT5 HTML │
│ ─► registry/gold_scalper_pro_xauusd_*_f{idx}.md │
│ （9 节自文档化 markdown，含差距表） │
└──────────────────────────────────────────────────────┘
```
--- ---
## What you must bring yourself ## 5. 常用脚本速查（按场景）
This KB is deliberately empty of trading IP. To build a working lab you supply: | 场景 | 脚本 | 命令 |
|------|------|------|
- **Your MQL5 EA(s)** — compiled `.ex5` plus source `.mq5`, and any custom indicators they call. | **跑 Optuna 搜索** | `optimize.py` | `--smoke`（30 trials，<2 分钟） / `--trials 500`（全量） |
- **Your strategy logic** — encoded once in Python (the caller) so the mirror engine can run it. | **重跑 finalist** | `reeval_finalist_forward.py` | 用 finalist 参数在 IS+OOS 重跑 Python，落 JSON |
- **Your presets** — the `.set` files / input templates you want to test and optimize. | **走前验证（forward）** | `walk_forward.py` | finalist 在 OOS 窗口独立验证 |
- **Historical data** — downloaded from your broker via MT5 (the stack includes a recipe). | **生成 dashboard** | `build_optuna_dashboard.py` | 中文 plotly HTML（5 主图 + 18 单参数 + 15 等高线） |
- **A broker demo account** — for the MT5 Strategy Tester runs. | **生成 ML 特征** | `build_feature_datasets.py` | trade-level 35 列 + trial-level parquet |
| **生成 registry** | `build_registry_entry.py` | 命令驱动生成 finalist 条目（**禁手写**） |
Everything else — the architecture, the optimizer, the MT5 bridge, the robustness checks, and the | **准备 MT5 验证** | `prepare_mt5_verify.py` | 生成 `.set` + `.ini`，复制到 MT5 tester profiles |
rules that hold it together — is described in the docs above. | **比对 finalist** | `compare_finalist.py` | Python vs MT5 差距表 |
| **诊断（9 个）** | `diag_*.py` | ATR / 时区 / 引擎 trace / 信号触发 / 仓位 mismatch 等 |
| **查 study** | `inspect_study.py` | 打印 trials / best params |
| **查 MT5 报告** | `inspect_report.py` / `diag_mt5_summary.py` / `diag_mt5_trades.py` | 解析 HTML |
| **下历史数据** | `download_xauusd_history.py`（M5）/ `download_xauusd_m1.py`（M1） | 必须 MT5 终端已登录 |
| **查 symbol 规格** | `query_xauusd_spec.py` | tick_value / contract_size / lot_step |
--- ---
*This KB is brand-free and self-contained. Drop the folder into a Git repository, open it with your ## 6. 当前项目状态
AI coding assistant, and build your own lab.*
### 6.1 已完成的 Phase
| Phase | 内容 | 状态 |
|-------|------|------|
| 0–1 | 设备 profile + Python 栈安装 | ✅ |
| 2 | 仓库骨架（shared/ 9 子包 + strategies/） | ✅ |
| 3 | MT5 连接（分步 initialize+login）+ EA 资产导入 | ✅ |
| 4 | ScalperEngine + M1 tick-level 模拟路径 | ✅ |
| 5 | XAUUSD_REAL instrument + GoldScalperPro search space | ✅ |
| 6 | Optuna 500-trial 搜索 + 3 diverse finalist | ✅ |
| 7 | MT5 forward mode IS/OOS 验证（旧 finalist）+ 差距根因 | ✅ |
| 8 | APPROVAL + PROMOTE（registry 第一条） | ✅ |
| — | Optuna warmup bug 修复（2026-06-26 重跑） | ✅ |
### 6.2 当前 finalist（warmup 修复后，2026-06-27 重跑）
来自 [studies/finalists/gold_scalper_pro_is2025-2026.json](studies/finalists/gold_scalper_pro_is2025-2026.json)：
| # | trial | score | IS net | OOS net | registry |
|---|-------|------:|-------:|--------:|----------|
| 1 | #491 | 186,075 | $263,527 | $10,632 | [f1_t491.md](registry/gold_scalper_pro_xauusd_2025-01-01_2026-06-26_f1_t491.md) |
| 2 | #297 | 68,772 | $94,197 | $11,153 | [f2_t297.md](registry/gold_scalper_pro_xauusd_2025-01-01_2026-06-26_f2_t297.md) |
| 3 | #492 | 129,575 | $165,740 | $6,969 | [f3_t492.md](registry/gold_scalper_pro_xauusd_2025-01-01_2026-06-26_f3_t492.md) |
> 这三个 finalist **尚未通过 MT5 验证**——registry 条目标记为
> "已记录（无 MT5 验证 — 待 Phase 7 MT5 验证）"。需要先在 MT5 Strategy Tester
> 跑出 IS+OOS HTML 报告，再用 `build_registry_entry.py --mt5-is-html ... --mt5-oos-html ...`
> 补齐 MT5 部分和差距表。
---
## 7. 接手者必读（续做指南）
### 7.1 三条铁律（违反任一条等于把项目搞烂）
1. **trailing/BE EA 必须传 `m1_bars=` 给 `engine.run()`**——否则 net profit
会有 −40% 到 −50% 的隐藏 gap。这不是噪声，是 bug。详见
[03-engine-design.md §7](03-engine-design.md)。
2. **`ObjectiveConfig` 必须传 `signals_full_bars=full_m5`**——否则 EMA/RSI/ATR
在 IS_START 才开始预热，首笔交易会比 MT5 晚 14 小时，整个 finalist 集合都
是错的。详见 [scripts/optimize.py](scripts/optimize.py) 的
`make_objective_config()` docstring。
3. **Engine 一旦验证就冻结，要试新想法就 fork**——直接改 validated engine
可能让所有已信任的数字都失效。Fork + default-OFF 实验 hook + 证明
fork-with-change-off == 原版 1:1 + 再 A/B。详见
[04-isolation-rules.md Rule 2](04-isolation-rules.md)。
### 7.2 已知坑（接手时大概率还会遇到）
| 坑 | 症状 | 修复 |
|----|------|------|
| MT5 连接 IPC 超时（-10005） | `mt5.initialize()` 直接传所有参数 | **分步**：先 `mt5.initialize(path=...)`，再 `mt5.login(login, password, server)` |
| 终端路径反斜杠 | IPC 超时 / 连不上 | 路径必须用**正斜杠** `/`，不是 `\` |
| MT5 进程残留 | 新连接失败 | 任务管理器清掉所有 `terminal64.exe` 进程后重试 |
| `.set` 文件加载失败 | MT5 tester 找不到参数 | 文件必须以 **BOM 开头**（`write_set_file` 已处理） |
| Plotly 图表看不见 | HTML 渲染 0 高度 | `fig.update_layout(height=<像素>)` + CDN + 单列 flex（见 `build_optuna_dashboard.py`） |
| 参数切片图标签重叠 | 5400px 宽复合图挤一起 | 拆成每参数独立 900×420 小图（`render_slice_grid`） |
| Optuna warmup bug | IS 起点附近 14h 没信号 | `signals_full_bars` 字段，详见 §7.1 第 2 条 |
### 7.3 急需做的下一步（按优先级）
1. **跑 MT5 验证新 finalist #1（trial #491）** —— 用
`scripts/prepare_mt5_verify.py` 生成 `.set` + `.ini`，在 MT5 Strategy Tester
forward mode 跑 IS=2025 + OOS=2026 H1，导 HTML 到 `reports/`，然后用
`build_registry_entry.py --mt5-is-html ... --mt5-oos-html ...` 补齐 registry。
2. **重新诊断 ATR 数据差异** —— 旧的 `diag_atr_check.py` / `diag_mt5_trades.py`
是针对 trial #324 的，新 finalist 需重跑。
3. **实现自动化 MT5 pipeline** —— `shared/mt5_pipeline/runner.py` 现在只是骨架，
把"打开 tester → 等跑完 → 拷 HTML"自动化掉。
4. **补全 robustness 层** —— `shared/robustness/layers.py` 只有
`stability_region`，缺 doc 06 §5 描述的其它层（parameter stability、
monte-carlo perturbation、walk-forward aggregate）。
5. **MT5 数据对齐** —— Python parquet M5 OHLC 与 MT5 tester 内部 history 在
IS 起点附近有微差，复利下被指数放大。需要决定是接受差距（旧 finalist
#324 已接受）还是修对齐。
### 7.4 不要做的事
- **不要手写 registry 条目**——必须用 `scripts/build_registry_entry.py` 生成。
脚本会从 finalist JSON + Optuna study + MT5 HTML 三源自动提取，包含
9 节内容（标识 / 参数表含百分位 / 三个 finalist 对比 / Python 指标 /
MT5 指标 / 差距分析 / 搜索统计 / 复现命令 / 已知限制）。
- **不要编辑已批准的 registry 条目**——append-only。新发现是新条目，
旧条目就算被超越也保留（负面结果也是数据）。
- **不要在 `engine.run()` 里加策略逻辑**——engine 只决定"价格是否触到
止损"，**从不决定止损放在哪**。换策略改 caller 和数组，engine 不动。
- **不要在 `mt5.initialize()` 里塞所有参数**——分步连接，详见 §7.2 第 1 条。
- **不要给单个 strategy import 另一个 strategy**——copy，不要 import。
iteration 各自拥有 snapshot，几个月前的 iteration 仍要能跑。
### 7.5 改动约定
- 改了 engine / objective / search space → 跑 `--smoke` 先验证全链路通
- 改了 finalist 挑选逻辑 → 重跑 `reeval_finalist_forward.py` + 重生 dashboard
- 改了 MT5 报告解析 → 用 `diag_mt5_summary.py` 对比新旧解析结果
- commit 时分小步、写清楚 why（不只是 what），git log 就是项目文档的一部分
---
## 8. 文档地图
| 你想做的事 | 读这个 |
|------------|--------|
| 跑一遍流程 | 本文件 §2 + §5 |
| 理解架构为什么这样搭 | [PROJECT_GUIDE.md §1](PROJECT_GUIDE.md) + [02-architecture.md](02-architecture.md) |
| 看具体踩过的坑 | [PROJECT_GUIDE.md §5](PROJECT_GUIDE.md) + 本文件 §7.2 |
| 写新策略 | [02-architecture.md §4](02-architecture.md) + [08-workflow-cycle.md](08-workflow-cycle.md) |
| 改 engine | [03-engine-design.md](03-engine-design.md) + [04-isolation-rules.md Rule 2](04-isolation-rules.md) |
| 改 Optuna objective | [06-optimization-and-robustness.md](06-optimization-and-robustness.md) + [shared/optimizer/objective.py](shared/optimizer/objective.py) |
| 接 MT5 | [07-mt5-bridge.md](07-mt5-bridge.md) + [shared/mt5_pipeline/](shared/mt5_pipeline/) |
| 看 finalist 参数 | [studies/finalists/gold_scalper_pro_is2025-2026.json](studies/finalists/gold_scalper_pro_is2025-2026.json) + [registry/](registry/) |
| 看 Optuna trials | `python scripts/inspect_study.py` 或直接打开 `studies/optuna/gold_scalper_pro_is2025.db` |
---
## 9. License & Disclaimer
个人研究项目，不构成投资建议。EA 源码 `GoldScalperPro.mq5` 属于其原作者；
本项目仅用于参数优化方法论研究，不重新分发 EA 本体。
_test_set_gen.py
-44
View File
@@ -1,44 +0,0 @@
"""Dry-run: generate .set + .ini with frozen baseline, verify format."""
import sys
from pathlib import Path
PROJECT = Path(__file__).resolve().parent
sys.path.insert(0, str(PROJECT))
from shared.config import get_secret, load_env
from shared.mt5_pipeline.ini_gen import TesterConfig, write_tester_ini, MODEL_OHLC
from shared.mt5_pipeline.set_gen import write_set_file
from strategies.gold_scalper_pro.search_space import FROZEN_BASELINE
from strategies.gold_scalper_pro.set_mappings import GOLD_SCALPER_MAPPINGS
load_env(PROJECT)
mt5_data = Path(get_secret("MT5_DATA_PATH"))
profiles = mt5_data / "MQL5" / "Profiles" / "Tester"
profiles.mkdir(parents=True, exist_ok=True)
set_path = profiles / "GoldScalperPro_dryrun.set"
write_set_file(FROZEN_BASELINE, GOLD_SCALPER_MAPPINGS, set_path)
print(f"set written: {set_path}")
print(f" size: {set_path.stat().st_size} bytes")
print(f" first bytes: {set_path.read_bytes()[:40]}")
ini_path = profiles / "GoldScalperPro_dryrun.ini"
tcfg = TesterConfig(
expert=r"Experts\GoldScalperPro.ex5",
symbol="XAUUSD",
period="M5",
model=MODEL_OHLC,
from_date="2024.06.26",
to_date="2026.06.26",
deposit=10000.0,
leverage=100,
report="report_dryrun",
shutdown_terminal=True,
set_file=str(set_path),
login=int(get_secret("MT5_DEMO_LOGIN")),
password=get_secret("MT5_DEMO_PASSWORD"),
server=get_secret("MT5_DEMO_SERVER"),
)
write_tester_ini(tcfg, ini_path)
print(f"\nini written: {ini_path}")
print(f"--- contents ---")
print(ini_path.read_text(encoding="utf-8"))
requirements.txt
+23
View File
@@ -0,0 +1,23 @@
# ── Core data stack ────────────────────────────────────────────────
pandas>=2.0 # DataFrame everywhere; bars/trades/metrics
numpy>=1.24 # vectorized indicator math
pyarrow>=14.0 # Parquet backend for pandas (.parquet bars & features)
# ── Optimizer ──────────────────────────────────────────────────────
optuna>=3.5 # TPE sampler, SQLite-persisted studies, visualization
sqlalchemy>=2.0 # Optuna's sqlite:// storage backend
plotly>=5.18 # optuna.visualization + dashboard HTML rendering
tqdm>=4.65 # progress bars (also used by optuna)
# ── MT5 bridge (Windows-only) ─────────────────────────────────────
MetaTrader5>=5.0.45 # connect MT5 terminal, pull history, query symbol spec
# NOTE: only installs on Windows. On non-Windows dev
# boxes you can still edit shared/ + strategies/ code,
# but the data-download / verify scripts won't run.
# ── Report parsing & config ────────────────────────────────────────
lxml>=4.9 # parse MT5 Strategy Tester HTML reports (fast path)
html5lib>=1.1 # fallback HTML parser for malformed reports
PyYAML>=6.0 # shared/wizard/wizard.py — wizard-answers.yaml I/O
python-dotenv>=1.0 # optional .env loader (shared/config.py has a plain
# fallback, so this is convenience not a hard dep)
run.py
-153
View File
@@ -1,153 +0,0 @@
"""Phase 5+6 entry point: optimize GoldScalperPro on real XAUUSD bars.
Assembles ObjectiveConfig (engine + signals + search space + constraints),
runs Optuna, applies the diverse top-N selector, then runs the robustness
layers on the finalists and prints a report. This is the cycle that Phase 7
will feed into MT5 for verification.
Usage:
python run.py [--trials 200] [--top-n 3] [--deposit 10000]
"""
from __future__ import annotations
import argparse
import sys
import time
from pathlib import Path
PROJECT = Path(__file__).resolve().parent
sys.path.insert(0, str(PROJECT))
import optuna
from shared.core.engine import SizingInputs
from shared.data.loaders import load_bars
from shared.optimizer.objective import (
Constraints,
ObjectiveConfig,
build_objective,
)
from shared.optimizer.selector import select_diverse_topn
from shared.robustness.layers import stability_region
from strategies.gold_scalper_pro.instruments import XAUUSD_REAL
from strategies.gold_scalper_pro.scalper_engine import (
ScalperConfig,
ScalperEngine,
config_from_params,
engine_kwargs_from_params,
)
from strategies.gold_scalper_pro.search_space import (
FROZEN_BASELINE,
INT_PARAMS,
SEARCH_SPACE,
)
from strategies.gold_scalper_pro.signals import build_signals
def find_bars_file() -> Path:
"""Auto-find the latest XAUUSD M5 parquet in data/."""
data_dir = PROJECT / "data"
candidates = sorted(data_dir.glob("XAUUSD_M5_*.parquet"))
if not candidates:
raise FileNotFoundError(f"no XAUUSD M5 parquet in {data_dir}")
return candidates[-1]
def main() -> int:
ap = argparse.ArgumentParser(description="Optimize GoldScalperPro.")
ap.add_argument("--trials", type=int, default=100,
help="Optuna trials (default 100)")
ap.add_argument("--top-n", type=int, default=3,
help="diverse finalists to select (default 3)")
ap.add_argument("--deposit", type=float, default=10000.0,
help="initial deposit (default 10000)")
ap.add_argument("--bars", type=str, default="",
help="path to parquet bars (blank = auto-find)")
args = ap.parse_args()
bars_path = Path(args.bars) if args.bars else find_bars_file()
print(f"=== GoldScalperPro optimization ===")
print(f"bars : {bars_path.name}")
print(f"trials : {args.trials}")
print(f"top-n : {args.top_n}")
print(f"deposit : {args.deposit:,.0f} USD")
print()
bars = load_bars(bars_path)
print(f"loaded {len(bars):,} bars {bars['timestamp'].iloc[0]} → {bars['timestamp'].iloc[-1]}")
# ── Assemble the objective ────────────────────────────────────────────
constraints = Constraints(
min_trades=25,
min_profit_factor=1.2, # relaxed for first pass; tightened later
max_equity_dd_pct=0.40, # 40% hard cap
)
obj_cfg = ObjectiveConfig(
engine=ScalperEngine(),
bars=bars,
instrument=XAUUSD_REAL,
sizing=SizingInputs(),
initial_deposit=args.deposit,
search_space=SEARCH_SPACE,
int_params=INT_PARAMS,
frozen_baseline=FROZEN_BASELINE,
constraints=constraints,
dd_weight=1.0,
build_signals=build_signals,
build_engine_kwargs=engine_kwargs_from_params,
)
objective = build_objective(obj_cfg)
# ── Run Optuna ─────────────────────────────────────────────────────────
optuna.logging.set_verbosity(optuna.logging.WARNING)
study = optuna.create_study(direction="maximize",
sampler=optuna.samplers.TPESampler(seed=42))
print(f"\nrunning {args.trials} trials ...")
t0 = time.time()
study.optimize(objective, n_trials=args.trials, show_progress_bar=False)
elapsed = time.time() - t0
print(f"done in {elapsed:.1f}s ({elapsed/args.trials:.2f}s/trial)")
# ── Report ────────────────────────────────────────────────────────────
best = study.best_trial
print(f"\n=== best trial #{best.number} ===")
print(f" score : {best.value:+,.2f}")
print(f" net profit : {best.user_attrs['net_profit']:+,.2f}")
print(f" profit factor : {best.user_attrs['profit_factor']:.2f}")
print(f" trades : {best.user_attrs['total_trades']}")
print(f" win rate : {best.user_attrs['win_rate']:.2%}")
print(f" equity DD : {best.user_attrs['max_equity_dd']:,.2f} "
f"({best.user_attrs['max_equity_dd_pct']:.2%})")
print(f" sharpe : {best.user_attrs['sharpe']:.2f}")
if best.user_attrs.get("violations"):
print(f" violations : {best.user_attrs['violations']}")
print(" params:")
for k, v in best.user_attrs["params"].items():
if k in SEARCH_SPACE:
print(f" {k:24s} = {v}")
# ── Diverse top-N ─────────────────────────────────────────────────────
print(f"\n=== diverse top-{args.top_n} finalists ===")
finalists = select_diverse_topn(study, args.top_n, SEARCH_SPACE)
for i, t in enumerate(finalists, 1):
print(f" #{i} trial {t.number}: score={t.value:+,.2f} "
f"net={t.user_attrs['net_profit']:+,.2f} "
f"PF={t.user_attrs['profit_factor']:.2f} "
f"trades={t.user_attrs['total_trades']}")
# ── Stability region (is the best on a plateau?) ─────────────────────
print(f"\n=== stability region ===")
sr = stability_region(study, SEARCH_SPACE)
print(f" passed : {sr.get('passed')}")
print(f" cluster_size : {sr.get('cluster_size')}")
print(f" best_in_cluster : {sr.get('best_in_cluster')}")
if sr.get("reason"):
print(f" reason : {sr['reason']}")
print("\n=== done ===")
print("Next: Phase 7 — generate .set/.ini for each finalist, run MT5, compare.")
return 0
if __name__ == "__main__":
raise SystemExit(main())